INDEX

Explanations

terms related to gradients and scores in the context of evaluation metrics or algorithms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 gră

-0.74

 gioielli

-0.61

 belangrij

-0.59

 rojos

-0.57

 nicio

-0.56

 buone

-0.55

EClass

-0.55

 acoper

-0.54

 renovar

-0.54

 crainte

-0.54

POSITIVE LOGITS

 gradient

1.06

gradient

0.94

 Gradient

0.93

Gradient

0.90

 gradients

0.79

 Review

0.75

Pattern

0.72

verifyException

0.71

pattern

0.70

 VIDEOTAPE

0.67

Activations Density 0.122%

terms related to gradients and scores in the context of evaluation metrics or algorithms

No Comments

No Known Activations