INDEX

Explanations

specific parts or sections of something

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "portion" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ATHER

-0.69

Tok

-0.68

CV

-0.66

generic

-0.66

Phill

-0.64

Trend

-0.63

ampa

-0.62

husband

-0.61

 Fighters

-0.61

Pope

-0.61

POSITIVE LOGITS

 thereof

1.03

edly

0.89

icularly

0.86

 guiActiveUn

0.84

 portions

0.80

xual

0.77

mble

0.75

uality

0.74

ONSORED

0.72

 portion

0.72

Activations Density 0.012%

specific parts or sections of something

mentions of the word "portion" in various contexts

No Comments

No Known Activations