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

 foregoing

-0.70

imens

-0.67

%]

-0.66

geries

-0.65

 specimens

-0.62

 mete

-0.61

rily

-0.60

 unprepared

-0.59

osponsors

-0.59

 holdings

-0.59

POSITIVE LOGITS

RW

0.77

XT

0.77

abul

0.76

Pod

0.75

ANG

0.75

AD

0.72

Cast

0.72

PR

0.71

perture

0.71

DB

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.