INDEX

Explanations

legal terms, specifically related to court appeals

oai_token-act-pair · gpt-3.5-turbo

references to appellate courts

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

pload

-0.78

ku

-0.78

 Wrestle

-0.71

 Horus

-0.71

zl

-0.70

 combust

-0.66

kson

-0.62

 Serpent

-0.62

shaw

-0.62

 stead

-0.62

POSITIVE LOGITS

 Appeals

1.05

 Tribunal

0.97

 hearings

0.96

 tribunal

0.82

 commissions

0.80

 Court

0.80

Court

0.80

 Circuit

0.79

rator

0.78

istrate

0.77

Activations Density 0.010%

legal terms, specifically related to court appeals

references to appellate courts

No Comments

No Known Activations