INDEX

Explanations

pronouns or possessive adjectives followed by action verbs describing mental or physical states or actions

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns and their associated contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

avers

-0.76

enced

-0.69

 parted

-0.66

aiden

-0.65

pps

-0.64

ĸļ

-0.63

Plex

-0.63

wered

-0.62

 shorth

-0.62

ochet

-0.62

POSITIVE LOGITS

ikuman

0.74

 stairs

0.73

 conflic

0.71

 ILCS

0.66

 seams

0.66

 calories

0.65

 Ballistic

0.64

enegger

0.64

ISSION

0.63

gam

0.63

Activations Density 0.340%

pronouns or possessive adjectives followed by action verbs describing mental or physical states or actions

possessive pronouns and their associated contexts

No Comments

No Known Activations