INDEX

Explanations

expressions related to reckless behavior or actions

oai_token-act-pair · gpt-3.5-turbo

references to reckless behavior

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

ļéĨĴ

-0.88

olithic

-0.85

culosis

-0.77

*/(

-0.76

rose

-0.76

 Rite

-0.71

olith

-0.71

berman

-0.71

ophon

-0.70

quart

-0.70

POSITIVE LOGITS

 reck

0.89

 endanger

0.88

 disregard

0.85

lessly

0.81

err

0.79

 ACTIONS

0.78

 negligence

0.77

 careless

0.75

 reckless

0.74

respons

0.73

Activations Density 0.036%

expressions related to reckless behavior or actions

references to reckless behavior

No Comments

No Known Activations