INDEX

Explanations

references to specific individuals or annotations in a context

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

The token "obj" or "object"

np_acts-logits-general · gemini-2.0-flash

objectobj

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Pennington

-0.53

OutputStream

-0.51

 Hann

-0.47

 Leven

-0.47

AutoScaleMode

-0.46

Ārējās

-0.45

teros

-0.45

Filmographie

-0.44

 Matheson

-0.44

 psoriasis

-0.44

POSITIVE LOGITS

obj

1.14

Ob

1.11

 object

1.08

obj

1.05

ob

1.05

object

1.04

ab

1.04

 аб

1.03

Ob

1.03

 OBJECT

1.03

Activations Density 0.247%

references to specific individuals or annotations in a context

The token "obj" or "object"

objectobj

No Comments

No Known Activations