INDEX

Explanations

mentions of hierarchical structures or levels within organizations

oai_token-act-pair · gpt-3.5-turbo

phrases related to hierarchical classifications or positions within groups

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

argo

-0.80

issance

-0.77

 Robo

-0.66

Interstitial

-0.62

 Pilgrim

-0.61

obo

-0.60

\/\/

-0.59

uras

-0.59

 Archangel

-0.59

 Peel

-0.58

POSITIVE LOGITS

 ranks

1.02

 insign

0.88

 rank

0.81

chwitz

0.75

Rank

0.74

wagon

0.74

 ranked

0.74

etting

0.73

paces

0.72

chool

0.71

Activations Density 0.010%

mentions of hierarchical structures or levels within organizations

phrases related to hierarchical classifications or positions within groups

No Comments

No Known Activations