INDEX

Explanations

phrases related to jumping onto something

oai_token-act-pair · gpt-3.5-turbo

phrases and expressions related to the action of "jumping on" or participating in something

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

Policy

-0.89

Limited

-0.73

ãĥĩãĤ£

-0.67

lt

-0.65

lf

-0.65

foreign

-0.63

policy

-0.63

invoke

-0.60

Repl

-0.60

 Bild

-0.60

POSITIVE LOGITS

 hoops

0.93

obin

0.82

 heels

0.79

 leaps

0.77

 heights

0.77

 stairs

0.75

ulic

0.75

aciously

0.73

 sidx

0.72

ioxide

0.71

Activations Density 0.128%

phrases related to jumping onto something

phrases and expressions related to the action of "jumping on" or participating in something

No Comments

No Known Activations