INDEX

Explanations

sentences indicating strong emotion or emphasis

oai_token-act-pair · gpt-3.5-turbo

statements affirming or asserting something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

dies

-0.78

elve

-0.74

luaj

-0.74

asers

-0.72

esm

-0.70

ievers

-0.70

styles

-0.69

IDs

-0.69

anguages

-0.68

ses

-0.67

POSITIVE LOGITS

 happening

0.97

 definitely

0.92

 supposed

0.91

 unacceptable

0.89

NOT

0.88

 gonna

0.87

 shaping

0.85

 bullshit

0.82

 ridiculous

0.82

 hardly

0.80

Activations Density 0.110%

sentences indicating strong emotion or emphasis

statements affirming or asserting something

No Comments

No Known Activations