INDEX

Explanations

phrases indicating a resorting or turning to a specific action or behavior

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions of resorting or indulgence

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

workers

-0.80

facing

-0.74

 Guatem

-0.73

enery

-0.70

si

-0.69

houses

-0.69

roy

-0.66

house

-0.64

rats

-0.63

GW

-0.62

POSITIVE LOGITS

 caution

0.75

 sidx

0.73

ierre

0.72

agall

0.68

pmwiki

0.68

 ç¥ŀ

0.68

 Rule

0.64

 insanity

0.64

 compromise

0.62

uchin

0.62

Activations Density 0.184%

phrases indicating a resorting or turning to a specific action or behavior

phrases indicating actions of resorting or indulgence

No Comments

No Known Activations