INDEX

Explanations

words related to high activity or involvement in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to being "active."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

aneers

-0.80

AMI

-0.77

ageddon

-0.75

umbo

-0.75

unts

-0.74

yth

-0.74

OUGH

-0.72

ĸļ

-0.69

obyl

-0.68

ERO

-0.67

POSITIVE LOGITS

 active

0.92

 participation

0.89

 newsp

0.83

 duty

0.83

 ingredient

0.82

 Directory

0.78

ives

0.78

 involvement

0.77

 participant

0.76

 activity

0.75

Activations Density 0.010%

words related to high activity or involvement in various contexts

terms and phrases related to being "active."

No Comments

No Known Activations