INDEX

Explanations

words related to emotions

oai_token-act-pair · gpt-3.5-turbo

prefixes or syllable patterns

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

 Companion

-0.84

 Shepherd

-0.81

 Worlds

-0.80

 Chronicle

-0.78

 Healer

-0.77

Kop

-0.76

 Centauri

-0.76

 Explorer

-0.76

 Tone

-0.75

 Defenders

-0.75

POSITIVE LOGITS

anc

1.10

abor

0.97

aring

0.96

secut

0.95

ared

0.94

ardless

0.93

ast

0.92

withstanding

0.92

atri

0.91

ic

0.91

Activations Density 0.194%

words related to emotions

prefixes or syllable patterns

No Comments

No Known Activations