INDEX

Explanations

phrases related to removing or eliminating something

oai_token-act-pair · gpt-3.5-turbo

phrases expressing a desire to eliminate or remove something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

BE

-0.70

anwhile

-0.70

 Fallon

-0.68

 Immortal

-0.67

OPER

-0.64

 cited

-0.64

 Macro

-0.62

DT

-0.61

ULT

-0.61

 Glob

-0.59

POSITIVE LOGITS

itaire

0.99

rid

0.90

ª

0.89

ients

0.88

anium

0.88

brush

0.84

atives

0.80

ges

0.79

icide

0.78

orget

0.77

Activations Density 0.007%

phrases related to removing or eliminating something

phrases expressing a desire to eliminate or remove something

No Comments

No Known Activations