INDEX

Explanations

pronouns related to identity and reference, particularly "who" and "which"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 PhpStorm

-0.56

 Iberian

-0.52

aarrggbb

-0.51

ophosphate

-0.51

openha

-0.51

 Crocker

-0.49

AccessorTable

-0.48

czas

-0.47

orne

-0.47

gyz

-0.46

POSITIVE LOGITS

 οποίο

0.86

 WHICH

0.71

 Which

0.71

 which

0.71

 also

0.69

########.

0.68

 οποία

0.67

GraphicsUnit

0.64

SuppressLint

0.61

 lequel

0.60

Activations Density 0.218%

pronouns related to identity and reference, particularly "who" and "which"

No Comments

No Known Activations