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

rito

-0.69

 Heights

-0.68

hetti

-0.65

ukemia

-0.65

livion

-0.65

peria

-0.64

;;;;;;;;;;;;

-0.64

————————

-0.62

 Till

-0.61

 spac

-0.61

POSITIVE LOGITS

uki

0.76

akuya

0.69

acles

0.68

aku

0.65

spir

0.65

 swept

0.64

ORPG

0.64

CHAT

0.61

RAL

0.61

aman

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.