INDEX

Explanations

proper nouns connected to locations, individuals, or organizations

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names or titles

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

ãĥŁ

-0.75

Reviewer

-0.72

SpaceEngineers

-0.70

ãģ¾

-0.66

 Cosponsors

-0.65

ENCY

-0.64

Glass

-0.60

SAN

-0.60

 srfAttach

-0.59

Preview

-0.58

POSITIVE LOGITS

alus

0.74

lict

0.67

hyde

0.67

utive

0.67

mort

0.66

liction

0.66

geon

0.65

 Restaur

0.65

ruction

0.65

arb

0.64

Activations Density 0.329%

proper nouns connected to locations, individuals, or organizations

proper nouns, particularly names or titles

No Comments

No Known Activations