INDEX

Explanations

terms related to evil or negative actions

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

licula

-0.53

stackrel

-0.46

ViewFeatures

-0.45

oples

-0.45

RT

-0.45

postIndex

-0.44

colhead

-0.44

Share

-0.44

DOT

-0.43

AutoScaleMode

-0.43

POSITIVE LOGITS

 evil

0.93

 predecess

0.89

 Evil

0.84

Evil

0.83

 Bucure

0.83

 costumi

0.78

 Româ

0.78

 Minang

0.78

 hilux

0.77

evil

0.76

Activations Density 0.092%

terms related to evil or negative actions

No Comments

No Known Activations