INDEX

Explanations

descriptive terms related to levels or qualities of something

oai_token-act-pair · gpt-3.5-turbo

terms associated with quality or rank

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

omen

-0.85

omal

-0.76

perty

-0.74

isl

-0.74

 unres

-0.72

TPPStreamerBot

-0.72

runtime

-0.71

otyp

-0.71

omers

-0.70

akis

-0.70

POSITIVE LOGITS

 notch

1.10

bler

0.75

chery

0.74

 buds

0.74

ments

0.72

birds

0.72

bird

0.70

IELD

0.70

!/

0.70

TPS

0.69

Activations Density 0.016%

descriptive terms related to levels or qualities of something

terms associated with quality or rank

No Comments

No Known Activations