INDEX

Explanations

phrases related to being actively engaged or participating in a situation or activity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "involved" in various contexts related to participation or engagement

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

mares

-0.75

OIL

-0.73

flu

-0.66

peat

-0.65

 passage

-0.63

mare

-0.63

 Bulg

-0.63

TI

-0.63

ipe

-0.63

omach

-0.63

POSITIVE LOGITS

ioned

0.84

enza

0.79

itaire

0.78

ulkan

0.75

Offline

0.74

cius

0.73

 involved

0.73

olved

0.72

ive

0.71

 actively

0.69

Activations Density 0.022%

phrases related to being actively engaged or participating in a situation or activity

instances of the word "involved" in various contexts related to participation or engagement

No Comments

No Known Activations