INDEX

Explanations

occurrences of character names and their actions within a narrative

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

states after events

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Autoritní

-0.43

 GONE

-0.38

 infeliz

-0.38

orated

-0.37

 echar

-0.37

 لكرة

-0.36

misa

-0.36

ärm

-0.35

 Crosse

-0.35

 Morte

-0.35

POSITIVE LOGITS

 recovery

0.59

 Recovery

0.52

 recoveries

0.49

Recovery

0.49

recovery

0.46

Recover

0.45

mybatisplus

0.45

 rekon

0.45

 heridas

0.44

Recovered

0.44

Activations Density 0.071%

occurrences of character names and their actions within a narrative

states after events

No Comments

No Known Activations