INDEX

Explanations

No Explanations Found

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

hani

-0.72

hm

-0.69

SN

-0.66

olog

-0.65

iru

-0.64

 Hindus

-0.62

Rs

-0.61

aro

-0.61

oi

-0.60

rily

-0.59

POSITIVE LOGITS

theless

0.81

ipop

0.76

abwe

0.74

helm

0.73

 metic

0.71

cue

0.70

 coma

0.69

isphere

0.68

uria

0.65

borgh

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.