INDEX

Explanations

statements related to personal opinions or declarations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

luaj

-0.78

opez

-0.78

ngth

-0.73

elve

-0.73

styles

-0.70

dies

-0.70

aneers

-0.69

alties

-0.67

anas

-0.66

ãĤ¨ãĥ«

-0.66

POSITIVE LOGITS

 happening

0.93

 true

0.75

 unacceptable

0.73

 nonsense

0.71

 untrue

0.71

 textbook

0.70

 HUGE

0.69

 purely

0.67

 blat

0.67

 assuming

0.65

Activations Density 0.758%

statements related to personal opinions or declarations

No Comments

No Known Activations