INDEX

Explanations

phrases related to causality and consequences

oai_token-act-pair · gpt-3.5-turbo

phrases indicating progression or continuity in a narrative

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

ullah

-0.65

essor

-0.64

ortium

-0.64

ament

-0.64

alis

-0.61

uminati

-0.61

ipation

-0.60

Interested

-0.60

owered

-0.59

urated

-0.58

POSITIVE LOGITS

verning

1.06

lems

0.91

ggle

0.87

vt

0.80

 overboard

0.80

along

0.79

©¶æ

0.79

 forth

0.75

 Forth

0.73

 Yards

0.73

Activations Density 0.091%

phrases related to causality and consequences

phrases indicating progression or continuity in a narrative

No Comments

No Known Activations