INDEX

Explanations

terms related to forensic analysis, particularly involving hair and starch

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

libft

-0.44

 برانيه

-0.43

tymologie

-0.43

</caption>

-0.42

 engraçadas

-0.40

AnchorStyles

-0.40

 plegable

-0.39

 eikä

-0.39

 productivo

-0.39

UnusedPrivate

-0.39

POSITIVE LOGITS

 forensic

0.70

 Forensic

0.61

Forensic

0.58

 criminal

0.54

 traces

0.46

 trace

0.42

 Crimin

0.42

 fore

0.41

 Autorizaciones

0.41

 Traces

0.40

Activations Density 0.125%

terms related to forensic analysis, particularly involving hair and starch

No Comments

No Known Activations