INDEX

Explanations

phrases indicating important consequences or impacts

oai_token-act-pair · gpt-3.5-turbo

statements about consequences or effects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fighters

-0.75

MQ

-0.72

cop

-0.70

cker

-0.68

fred

-0.68

bug

-0.66

few

-0.66

ced

-0.66

bows

-0.66

VERSION

-0.65

POSITIVE LOGITS

 implications

0.89

 beyond

0.84

romeda

0.82

 ramifications

0.81

ogene

0.76

uality

0.71

 arising

0.71

afety

0.69

 ripple

0.68

 consequential

0.67

Activations Density 0.053%

phrases indicating important consequences or impacts

statements about consequences or effects

No Comments

No Known Activations