INDEX

Explanations

locations and organizations mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

references to specific places and terms related to events or locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

loading

-0.82

ynamic

-0.80

lift

-0.79

dash

-0.76

wind

-0.74

âĸ¬âĸ¬

-0.73

price

-0.73

////////////////////////////////

-0.71

ãĥĨ

-0.69

less

-0.69

POSITIVE LOGITS

 Pompe

1.53

ername

0.87

ocalypse

0.84

cano

0.82

cale

0.80

uria

0.79

elong

0.75

uries

0.74

 externalToEVAOnly

0.73

ogn

0.72

Activations Density 0.012%

locations and organizations mentioned in the text

references to specific places and terms related to events or locations

No Comments

No Known Activations