INDEX

Explanations

text related to action figures or collectible figurines

oai_token-act-pair · gpt-3.5-turbo

mentions of action figures or collectibles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

velength

-0.79

é¾

-0.77

tnc

-0.76

 rainy

-0.73

renheit

-0.72

termination

-0.72

bey

-0.71

jamin

-0.71

ãĥĨãĤ£

-0.68

veyard

-0.67

POSITIVE LOGITS

figure

0.86

 figur

0.85

 prominently

0.80

 sculpt

0.77

 skating

0.76

 figure

0.76

 statues

0.76

 wraps

0.74

 sculptures

0.73

 figures

0.72

Activations Density 0.023%

text related to action figures or collectible figurines

mentions of action figures or collectibles

No Comments

No Known Activations