INDEX

Explanations

mentions of television or film series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nai

-0.72

usra

-0.71

prus

-0.70

stadt

-0.69

selves

-0.69

ibrary

-0.68

zie

-0.67

ividual

-0.67

zan

-0.66

bell

-0.65

POSITIVE LOGITS

 finale

1.08

 regulars

0.96

manship

0.87

 premiere

0.78

0.76

 opener

0.74

hew

0.71

enegger

0.71

hews

0.70

ional

0.70

Activations Density 0.150%

mentions of television or film series

No Comments

No Known Activations