INDEX

Explanations

phrases related to individuals being part of a specific group or organization

oai_token-act-pair · gpt-3.5-turbo

references to membership in organizations or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 relate

-0.81

 depiction

-0.72

 stink

-0.71

acs

-0.71

 explan

-0.68

initions

-0.68

 candles

-0.67

 nodd

-0.65

 traged

-0.65

ctions

-0.64

POSITIVE LOGITS

 Parliament

0.89

 Nanto

0.88

 parliament

0.88

 Highlands

0.85

 Physicians

0.83

 sorts

0.75

 Peaks

0.74

 UNESCO

0.74

 Population

0.72

 Surv

0.72

Activations Density 0.146%

phrases related to individuals being part of a specific group or organization

references to membership in organizations or groups

No Comments

No Known Activations