INDEX

Explanations

instances of terms related to document validity or invalidity

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 shenan

-0.67

 unspeak

-0.61

 impractica

-0.58

 unlaw

-0.57

 ecru

-0.54

 despotism

-0.54

 disagre

-0.54

 shuddered

-0.53

 indestru

-0.52

rası

-0.52

POSITIVE LOGITS

 valid

0.99

 Valid

0.97

 invalid

0.97

 VALID

0.96

Valid

0.93

 Invalid

0.93

valid

0.93

 validation

0.87

VALID

0.83

validation

0.83

Activations Density 0.070%

instances of terms related to document validity or invalidity

No Comments

No Known Activations