INDEX

Explanations

brand names and model identifiers associated with products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ãĥīãĥ©ãĤ´ãĥ³

-0.71

 captcha

-0.71

ãģį

-0.68

 Prosecut

-0.66

 Samoa

-0.65

oice

-0.65

#$

-0.65

Ú

-0.64

querade

-0.62

 Conversation

-0.61

POSITIVE LOGITS

unia

1.13

uran

1.02

mark

1.01

raph

0.97

lock

0.91

ula

0.89

mo

0.88

lar

0.84

Teg

0.83

eps

0.82

Activations Density 0.004%

brand names and model identifiers associated with products

No Comments

No Known Activations