INDEX

Explanations

words related to constructions or landmarks

oai_token-act-pair · gpt-3.5-turbo

various affixes and components common in scientific and technical terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Skydragon

-0.66

violence

-0.65

 riots

-0.63

LOD

-0.61

itiveness

-0.61

 immersion

-0.61

igation

-0.60

 disruptions

-0.60

 precincts

-0.60

quickShipAvailable

-0.59

POSITIVE LOGITS

ird

0.74

uld

0.74

ern

0.73

mann

0.72

ert

0.71

vre

0.71

 Pratt

0.70

rich

0.70

lich

0.69

irc

0.68

Activations Density 0.213%

words related to constructions or landmarks

various affixes and components common in scientific and technical terminology

No Comments

No Known Activations