INDEX

Explanations

specific mentions of city council members

oai_token-act-pair · gpt-3.5-turbo

instances of the word "council" and its variations

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

!/

-0.72

âĵĺ

-0.67

 Odyssey

-0.66

 Gamb

-0.65

 Faust

-0.63

 Achilles

-0.62

 Dawkins

-0.62

 Corinth

-0.61

 Patriot

-0.61

çīĪ

-0.61

POSITIVE LOGITS

cill

1.45

cil

1.05

 Coun

1.02

selling

0.92

plin

0.89

Coun

0.87

 councillor

0.81

itivity

0.79

 councillors

0.79

nai

0.79

Activations Density 0.011%

specific mentions of city council members

instances of the word "council" and its variations

No Comments

No Known Activations