INDEX

Explanations

words related to criminal offenses, specifically misdemeanors

oai_token-act-pair · gpt-3.5-turbo

terms related to legal classifications of offenses, particularly misdemeanors

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

rd

-0.78

riors

-0.78

arians

-0.76

ounding

-0.75

itious

-0.74

ative

-0.73

amar

-0.73

etics

-0.73

ulous

-0.70

ensibly

-0.69

POSITIVE LOGITS

mson

0.77

theless

0.74

cules

0.74

ught

0.70

sers

0.68

byss

0.68

 Trem

0.67

channelAvailability

0.67

BLE

0.67

ploy

0.66

Activations Density 0.062%

words related to criminal offenses, specifically misdemeanors

terms related to legal classifications of offenses, particularly misdemeanors

No Comments

No Known Activations