INDEX

Explanations

technical terms related to distinctions, differentiations, and classifications

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

 inconce

-0.65

 reluct

-0.60

 infallib

-0.53

 inappro

-0.50

 depic

-0.50

 insuper

-0.50

 impra

-0.50

 vehe

-0.49

 disagre

-0.49

 horrend

-0.49

POSITIVE LOGITS

 distinction

0.94

 distinguish

0.90

 distinctions

0.86

 DIST

0.83

 distinguishing

0.81

 Distinction

0.80

Distinguished

0.80

 distinguishes

0.80

 distinguished

0.79

distinct

0.77

Activations Density 0.082%

technical terms related to distinctions, differentiations, and classifications

No Comments

No Known Activations