INDEX

Explanations

doubt-related language, such as questioning and uncertainty expressed in text

oai_token-act-pair · gpt-3.5-turbo

statements about opinions or beliefs regarding societal issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

pleted

-0.76

 Tomorrow

-0.76

 Countdown

-0.75

 Caption

-0.71

 Next

-0.69

Pwr

-0.69

 Plans

-0.69

 Completed

-0.66

Tomorrow

-0.66

Next

-0.65

POSITIVE LOGITS

 empirical

1.00

 empir

0.98

 objectively

0.98

 intellectually

0.97

 philosophers

0.96

 normative

0.92

 logically

0.92

 morally

0.92

 patriarchy

0.91

 definitions

0.90

Activations Density 0.931%

doubt-related language, such as questioning and uncertainty expressed in text

statements about opinions or beliefs regarding societal issues

No Comments

No Known Activations