INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ratulations

-0.69

iciency

-0.66

 residues

-0.65

 independence

-0.65

 aggregation

-0.65

alions

-0.64

ng

-0.64

 vaccinations

-0.64

 replication

-0.64

etheus

-0.63

POSITIVE LOGITS

èĥ

0.73

#$#$

0.72

è¯

0.72

 Eater

0.69

edu

0.68

EStreamFrame

0.67

 Minotaur

0.65

Ultra

0.64

 redes

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.