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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

indal

-0.70

rising

-0.68

ayson

-0.68

cases

-0.67

sequence

-0.65

ogan

-0.64

ibrary

-0.63

tails

-0.63

selves

-0.63

weight

-0.62

POSITIVE LOGITS

nton

0.69

 Quart

0.64

 Tunnel

0.62

 Rabb

0.60

Hex

0.59

iosyn

0.59

pex

0.59

 encount

0.58

 satell

0.58

MAX

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.