INDEX

Explanations

mentions of bias or biased behaviors

oai_token-act-pair · gpt-3.5-turbo

references to bias 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 @ 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

Interstitial

-1.05

export

-0.84

eu

-0.82

clamation

-0.80

si

-0.80

ership

-0.76

ft

-0.74

akings

-0.74

allic

-0.71

PATH

-0.71

POSITIVE LOGITS

 bias

1.42

 biases

1.40

 prejud

1.00

 biased

0.98

 prejudice

0.98

 prejudices

0.90

 skew

0.85

 inference

0.85

 guiActiveUn

0.80

 creeps

0.79

Activations Density 0.008%

mentions of bias or biased behaviors

references to bias in various contexts

No Comments

No Known Activations