INDEX

Explanations

information related to research studies and citations in academic writing

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

references to statistical or numerical data in scientific contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 booted

-0.63

 ware

-0.61

 luck

-0.59

 undermin

-0.58

 mant

-0.58

 neighb

-0.58

 unbeat

-0.58

 stolen

-0.57

 enchanted

-0.57

 suspic

-0.56

POSITIVE LOGITS

],[

1.18

–

1.13

â

1.00

âĢĲ

1.00

âĢ¢âĢ¢

0.94

reviewed

0.89

].

0.87

âĪĴ

0.86

Î²

0.86

 Studies

0.79

Activations Density 0.053%

information related to research studies and citations in academic writing

references to statistical or numerical data in scientific contexts

No Comments

No Known Activations