INDEX

Explanations

expressions indicating feelings or opinions

oai_token-act-pair · gpt-3.5-turbo

the word "about" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Peaks

-0.82

KC

-0.76

------------------------

-0.72

rang

-0.71

gallery

-0.69

ammy

-0.68

ensation

-0.68

hesis

-0.66

CHR

-0.66

igmatic

-0.66

POSITIVE LOGITS

 respecting

0.78

 whether

0.71

how

0.71

 donating

0.70

 halfway

0.69

 reforming

0.67

 migrating

0.64

 coerc

0.64

 ducks

0.64

 protecting

0.64

Activations Density 0.102%

expressions indicating feelings or opinions

the word "about" in various contexts

No Comments

No Known Activations