INDEX

Explanations

references to policy and its implications in various contexts

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

 Stain

-0.79

iculture

-0.71

Ñĭ

-0.69

ergy

-0.66

ingly

-0.65

 Cruise

-0.64

apy

-0.64

Sic

-0.64

orf

-0.62

vae

-0.61

POSITIVE LOGITS

making

1.22

 makers

1.17

maker

1.12

makers

1.08

holders

1.07

holder

1.05

 prescriptions

0.97

Mic

0.84

 stances

0.84

 maker

0.83

Activations Density 0.103%

references to policy and its implications in various contexts

No Comments

No Known Activations