INDEX

Explanations

mentions of a specific location, Capitol Hill

oai_token-act-pair · gpt-3.5-turbo

references to Capitol Hill

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

theless

-0.72

gger

-0.72

hari

-0.72

ãģ¦

-0.71

cause

-0.70

erer

-0.69

guided

-0.68

okia

-0.68

::::::::

-0.67

word

-0.67

POSITIVE LOGITS

itol

1.04

 Hill

1.03

 Capitol

0.85

 Heights

0.85

itates

0.78

 dome

0.78

adium

0.74

 dioxide

0.73

ratulations

0.72

 Records

0.71

Activations Density 0.010%

mentions of a specific location, Capitol Hill

references to Capitol Hill

No Comments

No Known Activations