INDEX

Explanations

specific Java programming constructs or language features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

__':

-1.18

 nakalista

-1.06

 Emery

-0.99

 Majefty

-0.98

 purpoſe

-0.96

 Monfieur

-0.95

 "..\..\

-0.95

 pleaſure

-0.95

 onCancelled

-0.95

 "..\..\..\

-0.94

POSITIVE LOGITS

lang

1.75

 Lang

1.58

 lang

1.38

Lang

1.32

 LANG

1.26

 Langley

1.08

LANG

1.04

langs

0.95

ilang

0.78

 Langen

0.74

Activations Density 0.088%

specific Java programming constructs or language features

No Comments

No Known Activations