INDEX

Explanations

terms related to combat or military operations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "combat."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Else

-0.88

gow

-0.73

eus

-0.70

STER

-0.64

YE

-0.63

ULE

-0.62

Beck

-0.62

eton

-0.62

ocument

-0.62

Adds

-0.62

POSITIVE LOGITS

fighting

0.85

 combat

0.85

iveness

0.83

 fatig

0.82

combat

0.79

ments

0.79

ant

0.77

ants

0.73

fighters

0.73

stic

0.73

Activations Density 0.015%

terms related to combat or military operations

occurrences of the word "combat."

No Comments

No Known Activations