INDEX

Explanations

references to specific media outlets and platforms

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

 Television

-0.78

 à¨

-0.76

 Pipeline

-0.74

 Calculator

-0.73

 Scheme

-0.72

 Partial

-0.72

 Reconstruction

-0.72

 Examination

-0.72

 Ended

-0.72

 Directive

-0.71

POSITIVE LOGITS

yond

1.22

oft

1.12

bucks

1.08

blue

1.02

yp

1.02

orig

0.91

license

0.91

bad

0.91

reon

0.90

hesda

0.90

Activations Density 0.382%

references to specific media outlets and platforms

No Comments

No Known Activations