INDEX

Explanations

phrases related to serving or being in a position of service

oai_token-act-pair · gpt-3.5-turbo

references to military service and incarceration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 tackling

-0.73

isms

-0.64

 Collider

-0.63

Winged

-0.62

TPPStreamerBot

-0.62

 stumbling

-0.61

Jump

-0.60

Bridge

-0.60

tracking

-0.60

Weapon

-0.59

POSITIVE LOGITS

 purpose

0.84

 faithfully

0.80

 purposes

0.78

 subpoena

0.75

 constituents

0.74

 summons

0.72

olit

0.72

elli

0.71

 subpoen

0.70

 needs

0.68

Activations Density 0.120%

phrases related to serving or being in a position of service

references to military service and incarceration

No Comments

No Known Activations