INDEX

Explanations

concepts related to discrepancies and imbalances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

DebuggerNonUser

-0.49

ennio

-0.49

')));

-0.46

Proof

-0.44

fiques

-0.43

 Audiodateien

-0.43

']));

-0.43

 force

-0.42

 drop

-0.42

 proven

-0.41

POSITIVE LOGITS

 imbalance

1.15

 disambiguazione

1.04

 imbalances

1.03

 unbalanced

1.01

 uneven

0.99

 mismatch

0.94

 asymmetry

0.90

 discrepancy

0.89

 asym

0.84

 skewed

0.84

Activations Density 1.063%

concepts related to discrepancies and imbalances

No Comments

No Known Activations