INDEX

Explanations

phrases related to strong opinions or conclusions

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly periods and question marks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

itionally

-0.86

 intention

-0.78

 carbohyd

-0.74

 earthqu

-0.73

 advanced

-0.72

 teasp

-0.72

 intended

-0.71

 unlawfully

-0.71

isbury

-0.70

 interpreter

-0.70

POSITIVE LOGITS

 Anyway

1.90

Anyway

1.64

 Thankfully

1.36

 Luckily

1.34

 Nonetheless

1.34

 Needless

1.33

 Regardless

1.31

 Maybe

1.31

Hmm

1.30

 Nope

1.29

Activations Density 0.599%

phrases related to strong opinions or conclusions

punctuation marks, particularly periods and question marks

No Comments

No Known Activations