INDEX

Explanations

occurrences of the word "had"

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

woordig

-0.75

 pist

-0.70

Sammy

-0.69

orion

-0.67

SLS

-0.66

 Olsson

-0.65

olor

-0.64

 Torrey

-0.64

 Reuter

-0.64

PreferredItem

-0.64

POSITIVE LOGITS

 hoped

1.16

 earlier

1.05

HAD

1.02

Earlier

0.99

had

0.97

Had

0.97

 Earlier

0.92

 hadn

0.90

Had

0.90

 originalmente

0.87

Activations Density 0.153%

occurrences of the word "had"

No Comments

No Known Activations