INDEX

Explanations

phrases related to time durations or historical periods

oai_token-act-pair · gpt-3.5-turbo

mentions of time-related terms, particularly years and months

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

UL

-0.66

-0.59

 Valhalla

-0.58

 Elys

-0.54

 pecul

-0.54

 Cards

-0.53

 Pitch

-0.52

 Byte

-0.52

};

-0.51

BS

-0.51

POSITIVE LOGITS

izo

0.82

ordering

0.80

azard

0.75

eks

0.74

aid

0.71

 toward

0.71

abbling

0.70

essage

0.69

aird

0.68

thren

0.68

Activations Density 0.760%

phrases related to time durations or historical periods

mentions of time-related terms, particularly years and months

No Comments

No Known Activations