INDEX

Explanations

mentions of locations, with a focus on the state of "New Hampshire."

oai_token-act-pair · gpt-3.5-turbo

references to New Hampshire

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

SHIP

-0.83

ynt

-0.78

aic

-0.77

ovie

-0.71

henko

-0.69

zers

-0.67

ngth

-0.67

matically

-0.65

watching

-0.65

 behav

-0.64

POSITIVE LOGITS

 Hampshire

0.81

shire

0.76

 Wildcats

0.76

 Volunteers

0.74

 Pradesh

0.74

ers

0.72

 Divinity

0.71

 State

0.70

 Devils

0.68

idia

0.68

Activations Density 0.008%

mentions of locations, with a focus on the state of "New Hampshire."

references to New Hampshire

No Comments

No Known Activations