INDEX

Explanations

references to formal communication and written responses

oai_token-act-pair · gpt-3.5-turbo

references to journalism and media-related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.57

èĢ

-0.57

lust

-0.57

¥µ

-0.56

âĪĴ

-0.56

 unle

-0.52

 backward

-0.52

 intrinsically

-0.51

é¾įå

-0.51

efficient

-0.51

POSITIVE LOGITS

 emailed

0.96

 enqu

0.87

 contacting

0.84

 contacted

0.84

 email

0.77

 commenter

0.76

 info

0.76

 pics

0.74

 replies

0.74

 inquired

0.73

Activations Density 1.044%

references to formal communication and written responses

references to journalism and media-related topics

No Comments

No Known Activations