INDEX

Explanations

references to the word "gram" in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of "gram" and related terms, focusing on measurements and grammatical contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.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

ESCO

-0.67

orld

-0.66

Madison

-0.65

olve

-0.64

 Dying

-0.61

 Cust

-0.61

PsyNetMessage

-0.61

imaru

-0.61

 Arabian

-0.60

 Cind

-0.60

POSITIVE LOGITS

matical

1.55

matically

1.37

ophone

1.12

ming

1.01

mers

0.96

matic

0.96

bling

0.95

acy

0.89

bler

0.83

bled

0.82

Activations Density 0.042%

references to the word "gram" in various contexts

mentions of "gram" and related terms, focusing on measurements and grammatical contexts

No Comments

No Known Activations