INDEX

Explanations

words related to declarations or claims

oai_token-act-pair · gpt-3.5-turbo

references to declaration or classification terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

BIP

-0.71

 è£ı

-0.69

 tremend

-0.67

awaru

-0.65

ãĤ®

-0.64

ãĥ´ãĤ¡

-0.64

BALL

-0.63

Îº

-0.62

DEM

-0.61

Bey

-0.61

POSITIVE LOGITS

osures

1.09

avier

1.04

othes

1.04

osing

0.98

iff

0.98

ipper

0.97

inic

0.96

arent

0.95

utch

0.95

uster

0.94

Activations Density 0.012%

words related to declarations or claims

references to declaration or classification terms

No Comments

No Known Activations