INDEX

Explanations

words or phrases related to editing or revisions

oai_token-act-pair · gpt-3.5-turbo

information related to roles and classifications in various 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

ayson

-0.83

hement

-0.78

 neighb

-0.75

 Autob

-0.71

terday

-0.69

ividual

-0.68

 warr

-0.68

 citiz

-0.68

 ende

-0.68

undai

-0.67

POSITIVE LOGITS

References

0.93

Trivia

0.85

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.77

pmwiki

0.77

Joined

0.74

Effects

0.72

Soviet

0.72

Associated

0.71

Appearances

0.70

Gallery

0.69

Activations Density 0.090%

words or phrases related to editing or revisions

information related to roles and classifications in various contexts

No Comments

No Known Activations