INDEX

Explanations

statements or declarations made by entities or individuals

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "said" and related statements in a news context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

estern

-0.75

pes

-0.72

å¦

-0.72

cffffcc

-0.71

Friend

-0.69

odox

-0.67

ldom

-0.66

tnc

-0.66

zin

-0.66

é¾įå¥ĳå£«

-0.66

POSITIVE LOGITS

its

1.18

 itself

1.00

Its

0.87

 they

0.83

 yesterday

0.82

ITS

0.82

Its

0.81

 goodbye

0.78

 Wednesday

0.76

 Thursday

0.75

Activations Density 0.131%

statements or declarations made by entities or individuals

occurrences of the word "said" and related statements in a news context

No Comments

No Known Activations