INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Townsend

-0.68

ãĤ«

-0.66

burgh

-0.66

Edge

-0.59

 Methods

-0.58

 coolest

-0.58

idas

-0.58

 Patient

-0.58

Effect

-0.58

Iro

-0.57

POSITIVE LOGITS

vernment

0.93

iversal

0.77

MpServer

0.76

regor

0.72

umbn

0.72

 unlaw

0.70

ongyang

0.70

rieved

0.69

ilateral

0.68

ilater

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.