INDEX

Explanations

words related to reaching a final decision or reaching an agreement

oai_token-act-pair · gpt-3.5-turbo

statements about conclusions and opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

xtap

-0.80

quet

-0.60

quin

-0.60

iries

-0.57

erc

-0.57

quer

-0.57

illions

-0.54

egu

-0.53

 trailed

-0.53

agu

-0.52

POSITIVE LOGITS

 that

1.12

that

1.09

 THAT

0.87

 thats

0.85

 otherwise

0.83

 THEY

0.82

 they

0.78

there

0.77

 there

0.77

That

0.72

Activations Density 0.408%

words related to reaching a final decision or reaching an agreement

statements about conclusions and opinions

No Comments

No Known Activations