INDEX

Explanations

terms related to specific names or individuals

oai_token-act-pair · gpt-3.5-turbo

terms related to physical objects and names, particularly in narrative contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dies

-0.71

encia

-0.71

enza

-0.71

Fran

-0.67

fast

-0.66

 subtitles

-0.64

é¾įå

-0.63

HOME

-0.63

Republic

-0.61

ners

-0.61

POSITIVE LOGITS

 mound

1.01

uzzle

0.89

ossier

0.89

pload

0.89

mble

0.84

hammad

0.84

xual

0.81

resses

0.80

otin

0.79

ifully

0.79

Activations Density 0.015%

terms related to specific names or individuals

terms related to physical objects and names, particularly in narrative contexts

No Comments

No Known Activations