INDEX

Explanations

words related to news broadcasting

oai_token-act-pair · gpt-3.5-turbo

references to news broadcasts or reports

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

SR

-0.75

Seg

-0.74

DI

-0.65

MpServer

-0.65

isha

-0.63

BE

-0.62

EO

-0.61

SI

-0.61

 Ferr

-0.61

 Scal

-0.61

POSITIVE LOGITS

ansion

0.85

ģĸ

0.85

theless

0.81

urities

0.78

ateg

0.76

perature

0.75

sylvania

0.74

istration

0.74

akable

0.74

atown

0.73

Activations Density 0.043%

words related to news broadcasting

references to news broadcasts or reports

No Comments

No Known Activations