INDEX

Explanations

proper nouns or specific terms

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "known" indicating the introduction of a term or concept

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hibition

-0.65

rentice

-0.63

itiveness

-0.63

IRO

-0.60

 fixation

-0.56

olicy

-0.56

raining

-0.55

psey

-0.54

osion

-0.54

luence

-0.53

POSITIVE LOGITS

 collectively

1.07

as

0.97

 locally

0.92

 coll

0.90

 internally

0.89

 euphem

0.89

 abbrevi

0.87

 alternatively

0.86

 pseudonym

0.82

 simply

0.82

Activations Density 0.037%

proper nouns or specific terms

phrases that include the word "known" indicating the introduction of a term or concept

No Comments

No Known Activations