INDEX

Explanations

survey or poll respondent-related information

oai_token-act-pair · gpt-3.5-turbo

mentions of survey respondents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Horizon

-0.67

 Formation

-0.67

 Tours

-0.67

 Renaissance

-0.64

 Herald

-0.63

 enlarged

-0.62

 brakes

-0.62

Enhanced

-0.61

 Vengeance

-0.60

 Colonial

-0.59

POSITIVE LOGITS

 respondents

1.33

 respondent

1.07

 Respond

0.88

uggest

0.86

 undecided

0.85

 answ

0.83

swer

0.76

DK

0.74

ucker

0.74

ecided

0.72

Activations Density 0.014%

survey or poll respondent-related information

mentions of survey respondents

No Comments

No Known Activations