INDEX

Explanations

phrases related to behavior or traits that are typical or often observed in particular situations

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate habitual behaviors or inclinations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aban

-0.80

idden

-0.77

mberg

-0.75

gur

-0.75

ÄŁ

-0.73

estamp

-0.72

loo

-0.72

riel

-0.71

oho

-0.71

ighth

-0.71

POSITIVE LOGITS

 toward

1.05

 towards

0.99

 tendency

0.97

 tendencies

0.95

 predis

0.78

 shift

0.75

 propensity

0.75

 behavior

0.73

 Towards

0.72

 behaviour

0.72

Activations Density 0.023%

phrases related to behavior or traits that are typical or often observed in particular situations

phrases that indicate habitual behaviors or inclinations

No Comments

No Known Activations