INDEX

Explanations

references to instructions or steps in a list

oai_token-act-pair · gpt-3.5-turbo

occurrences of parentheses in the text

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

 confront

-0.79

 spir

-0.79

 constitu

-0.72

 prey

-0.71

 clin

-0.70

 grip

-0.70

 gang

-0.69

 detained

-0.69

 confronted

-0.68

 aver

-0.68

POSITIVE LOGITS

excluding

1.55

optional

1.50

depending

1.45

which

1.44

assuming

1.42

except

1.41

unless

1.41

usually

1.41

including

1.39

possibly

1.38

Activations Density 0.132%

references to instructions or steps in a list

occurrences of parentheses in the text

No Comments

No Known Activations