INDEX

Explanations

sequences indicating the presence of instructions or directives

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

mybatisplus

-1.07

awtextra

-1.04

 pinulongan

-0.86

CPtr

-0.86

Rüyada

-0.85

 Vikipedi

-0.83

protoimpl

-0.83

 disambiguazione

-0.82

HasForeignKey

-0.82

awaiter

-0.81

POSITIVE LOGITS

0.70

0.69

_.

0.67

<eos>

0.65

__.

0.59

{.

0.57

.$.

0.56

).

0.54

↵↵

0.53

庁

0.53

Activations Density 0.079%

sequences indicating the presence of instructions or directives

No Comments

No Known Activations