INDEX

Explanations

references to reading materials or exercises

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

-0.59

wußt

-0.46

 AssemblyCompany

-0.45

 plupart

-0.44

 uffic

-0.43

 Públicas

-0.42

 wuß

-0.42

 duiz

-0.42

 vroe

-0.42

 nyata

-0.42

POSITIVE LOGITS

 Reading

1.01

Reading

0.95

 reading

0.85

 READING

0.85

reading

0.84

 Readings

0.84

READING

0.80

Readings

0.79

 readings

0.76

 Reads

0.64

Activations Density 0.008%

references to reading materials or exercises

No Comments

No Known Activations