INDEX

Explanations

terms related to legal and judicial settings

oai_token-act-pair · gpt-3.5-turbo

concepts related to judgment or assessment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

AMP

-0.57

 Aden

-0.56

raft

-0.56

MAP

-0.55

yrinth

-0.55

ICES

-0.53

 Nepal

-0.53

 consolidation

-0.52

 Locations

-0.51

 Canaver

-0.51

POSITIVE LOGITS

?,

1.02

?),

0.93

?).

0.93

 lover

0.85

guy

0.81

 liar

0.81

 bigot

0.80

 idiot

0.80

 deity

0.79

>,

0.79

Activations Density 0.857%

terms related to legal and judicial settings

concepts related to judgment or assessment

No Comments

No Known Activations