INDEX

Explanations

reports or statements made by individuals

oai_token-act-pair · gpt-3.5-turbo

reports or statements made by officials or sources

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

 ILCS

-0.78

reci

-0.72

racted

-0.72

spir

-0.68

untled

-0.67

adesh

-0.64

MIT

-0.62

oppable

-0.61

Untitled

-0.61

Contents

-0.61

POSITIVE LOGITS

tale

0.99

mares

0.82

ingly

0.80

 reporters

0.79

us

0.78

warn

0.76

me

0.75

rich

0.73

IPS

0.73

ariat

0.73

Activations Density 0.060%

reports or statements made by individuals

reports or statements made by officials or sources

No Comments

No Known Activations