INDEX

Explanations

phrases related to legal imprisonment or sentencing

oai_token-act-pair · gpt-3.5-turbo

phrases related to criminal charges or sentencing

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.79

Tool

-0.77

edin

-0.75

ieth

-0.74

options

-0.74

ern

-0.73

atl

-0.73

ise

-0.70

ilant

-0.70

POSITIVE LOGITS

 violating

1.28

 breaching

1.20

 nonviolent

1.06

 failing

1.05

 violations

1.05

 treason

1.02

 crimes

1.01

 offences

1.01

 refusing

0.99

 felony

0.98

Activations Density 0.116%

phrases related to legal imprisonment or sentencing

phrases related to criminal charges or sentencing

No Comments

No Known Activations