INDEX

Explanations

words relating to names

oai_token-act-pair · gpt-3.5-turbo

human names or names of individuals related to a context

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

staking

-0.85

tails

-0.78

ebted

-0.76

cens

-0.73

 Scare

-0.73

 conduc

-0.71

ãģ®éŃĶ

-0.70

enegger

-0.69

tailed

-0.68

tail

-0.66

POSITIVE LOGITS

mn

0.76

ng

0.76

rs

0.76

ctr

0.75

xon

0.74

cs

0.72

mt

0.71

mr

0.71

alm

0.71

TG

0.69

Activations Density 0.192%

words relating to names

human names or names of individuals related to a context

No Comments

No Known Activations