INDEX

Explanations

contact information related to verification processes or legal proceedings

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

words starting with "Ver", especially forms of "verify" or "version".

oai_token-act-pair · claude-3-5-sonnet-20240620 Triggered by @bot

the word "verification" or its variations.

oai_token-act-pair · gemini-1.5-flash Triggered by @bot

the word "verification" or "ver" when it appears as the start of the word and capitalized (likely at the beginning of a sentence).

oai_token-act-pair · gemini-1.5-pro Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ollo

-0.79

urst

-0.75

»Ĵ

-0.70

ĩ

-0.69

asta

-0.69

ħĭ

-0.68

erer

-0.67

OWER

-0.65

utics

-0.65

etic

-0.65

POSITIVE LOGITS

gence

1.13

dict

1.12

ifiable

1.03

itas

1.02

izons

0.99

tex

0.99

gent

0.98

itable

0.95

ified

0.95

ifying

0.95

Activations Density 0.995%

contact information related to verification processes or legal proceedings

words starting with "Ver", especially forms of "verify" or "version".

the word "verification" or its variations.

the word "verification" or "ver" when it appears as the start of the word and capitalized (likely at the beginning of a sentence).

No Comments

No Known Activations