INDEX

Explanations

pronouns and verbs indicating belief or ability

oai_token-act-pair · gpt-3.5-turbo

pronouns indicating possession or agency

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

 suggestion

-0.68

wik

-0.64

focus

-0.60

summary

-0.60

 Broad

-0.59

 Bucks

-0.58

edia

-0.58

gard

-0.58

 Torrent

-0.57

 Fraz

-0.57

POSITIVE LOGITS

've

0.97

self

0.93

selves

0.90

'd

0.89

're

0.88

'll

0.85

 could

0.85

athered

0.84

areth

0.80

owe

0.78

Activations Density 0.172%

pronouns and verbs indicating belief or ability

pronouns indicating possession or agency

No Comments

No Known Activations