INDEX

Explanations

proper nouns, specifically names related to politics and leadership

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals or characters

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

 Whitman

-0.70

Fed

-0.68

Sed

-0.68

 Plum

-0.65

 Bullets

-0.65

Reviewer

-0.64

Spread

-0.64

REDACTED

-0.63

 Harriet

-0.63

 Sergeant

-0.63

POSITIVE LOGITS

emer

0.79

ymes

0.77

ippers

0.76

\\\\

0.76

agan

0.75

oÄŁ

0.74

ark

0.71

oos

0.70

pid

0.70

ille

0.69

Activations Density 0.032%

proper nouns, specifically names related to politics and leadership

mentions of specific individuals or characters

No Comments

No Known Activations