INDEX

Explanations

terms related to the "Motor" industry, such as company names and events

oai_token-act-pair · gpt-3.5-turbo

references to automotive manufacturers and motorsport contexts

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

ylum

-0.80

ilts

-0.77

 Foley

-0.74

xual

-0.71

wered

-0.71

 Ming

-0.68

yip

-0.68

 Ibid

-0.67

avorite

-0.67

tnc

-0.65

POSITIVE LOGITS

cycle

1.40

cycles

1.22

cycl

1.00

puter

0.94

cars

0.90

otor

0.85

bike

0.84

immune

0.82

wagen

0.80

company

0.77

Activations Density 0.012%

terms related to the "Motor" industry, such as company names and events

references to automotive manufacturers and motorsport contexts

No Comments

No Known Activations