INDEX

Explanations

negative terms related to trends or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

BibitemShut

-1.18

RectangleBorder

-1.12

 وتسجيلات

-1.10

 Paglinawan

-1.07

bibfield

-1.06

DockStyle

-1.06

Hauptartikel

-1.02

 Signalez

-1.00

 Италијани

-1.00

 AssemblyVersion

-0.98

POSITIVE LOGITS

0.77

<eos>

0.76

0.66

‑

0.64

0.63

‘

0.61

’

0.60

0.59

‐

0.58

Activations Density 0.306%

negative terms related to trends or processes

No Comments

No Known Activations